Eye-on-Hand Calibration Method for Cable-Driven Parallel Robots

International audienceEstimating the geometric parameters of a cable-driven parallel robot (CDPR) can be a labour intensive process or one that requires expensive sensors. This paper presents a low-cost method for estimating initial cable lengths and fixed cable attachment points of CDPRs. The proposed approach relies on the detection, mapping and localisation of fiduciary markers in the robot environment using a camera attached to the end effector. This paper additionally tackles the generation of a list of reachable calibration poses and presents a control scheme allowing the CDPR to reach those. Experiments are also carried out to assess the performance of the proposed calibration method. It appears that the proposed eye-on-hand method is more accurate than a previously reported method relying solely on cable-length measurements